Health care institutions and providers face mounting pressure to wring more value out of every dollar spent on caring for their patients.
A new review shows that most efforts to decrease low-value care have based their measurement of success on how much they reduced the overall use of certain tests and treatments. Far fewer looked at whether these efforts actually ensured that patients got more appropriate care and avoided unintended negative consequences.
The review, published in the Journal of General Internal Medicine, looks at 117 efforts aimed at reducing low-value care and how they measured the effects of these efforts.
“Low-value” can mean many things, including care that doesn’t benefit patients and could even harm them, wastes limited health care resources or leads to unnecessary costs.
Hundreds of studies over the past two decades have revealed many services that lack value for some or all patients. Patients and clinicians now have easy-to-follow guidance on what those are, thanks to the Choosing Wisely campaign from the American Board of Internal Medicine Foundation.
The new review focuses on what happens when teams act on this evidence and guidance, and researchers try to study the effects.
The bottom line? Those trying to reduce low-value care should take a bigger-picture view.
The authors, led by health care researchers from the University of Michigan, the VA Ann Arbor Center for Clinical Management Research and the University of Toronto, performed the review at the request of AcademyHealth, a non-profit professional society focused on improving health and health care by moving researchers’ evidence into action. The study was funded by the Patient-Centered Outcomes Research Institute.
“Reducing use of low-value services is important, but in doing so we need to also make sure we are assessing things that are clinically relevant, like whether appropriate care is being delivered to patients rather than only whether use of a given service is being reduced,” says Jennifer Maratt, M.D., a clinical lecturer in the U-M Department of Internal Medicine and the VA Ann Arbor Healthcare System who led the study with U-M colleagues Sameer Saini, M.D., and Eve Kerr, M.D.
More about the findings
The researchers looked at 101 papers published between 2010 to 2016 about specific efforts to reduce low-value care. They also examined 16 studies underway through ClinicalTrials.gov.
In all, 68 percent of the published efforts focused on measuring and changing the use of a particular test or treatment, but only 41 percent measured an outcome — that is, what happened when they changed that use. About half tried to gauge whether a test or treatment was appropriate for patients — arguably the most clinically meaningful measure.
But only one-third of these studies had looked for unintended consequences of their efforts to wring low-value care out of their care environment.
Such consequences — such as missing when an individual patient needs a certain treatment or test — can occasionally happen when an across-the-board cut in a particular medical service results in some patients not getting something that could have helped them specifically.
For instance, an effort to reduce overuse of antibiotics in hospitalized patients could unintentionally lead to more of them ending up at the emergency department later if an infection flares up.
“The Choosing Wisely campaign has dramatically increased the number of studies done to reduce low-value care, which is great,” says Kerr, a professor at U-M and director of the VA Center for Clinical Management Research. “However, we found that the majority of these studies do not assess outcomes that are truly meaningful to patients.”
A patient’s perspective
Most studies did not look for this kind of “backfiring,” and very few involved a patient perspective. In all, only 8 percent asked patients about the impact of the change on them — what researchers call a “patient-reported outcome.”
The 16 studies still in progress were a little better at aiming to take a big-picture view than the published studies.
Of these ongoing studies, 75 percent aim to measure a specific outcome of the effort, and 63 percent are looking for unintended consequences. Half include plans to measure patient-reported outcomes.
The researchers also found that ongoing studies are much more likely to use methods that meet the gold standard of research, including randomizing patients to a particular care group and using a control group for comparison.
Newer studies are also more likely to involve patients directly in efforts to reduce low-value care, mainly by educating them about whether a test or treatment is likely to benefit them.
Says Saini, “By focusing on simple utilization, the vast majority of studies provide an incomplete picture of the impact of these often powerful and complex interventions. For example, we often do not know how interventions to reduce use of low-value care affect the patient-provider relationship or to what extent they unintentionally lead to fewer tests or prescriptions in patients who need them.” Saini is an associate professor of medicine at U-M and research scientist at the VA Center for Clinical Management Research.
In general, the team says, researchers and evaluators should incorporate more clinically meaningful and patient-centered measures into studies to provide a more comprehensive understanding of the impact of these interventions.
The authors call for more standardization in how health care providers evaluate their efforts to reduce low-value care.
They also say more of these studies need to evaluate that the right services are being reduced in the right patients, that patient-provider relationships are assessed and that downstream outcomes improve.
Examples of interventions to reduce low-value care:
- Cost sharing and value-based purchasing
- Patient education and decision-making
- Quality indicators and reporting
- Physician performance incentives
- Utilization management
- Financial risk sharing/physician reimbursement
- Clinical decision support
- Provider education
- Provider feedback and peer reporting